Differential Biomarker Profiles Across Functional and Echocardiographic Phenotypes of Heart Failure
Keywords:
heart failure, NT-proBNP, vitamin D, hsCRP, NYHA, LVEFAbstract
Introduction: Heart failure is a heterogeneous clinical syndrome in which functional limitation and left ventricular systolic dysfunction may reflect distinct pathophysiological processes. Biomarkers of hemodynamic stress, metabolic status, and inflammation may show different associations with clinical and echocardiographic phenotypes.
The Aim: To investigate associations of NT-proBNP, vitamin D, and hsCRP with NYHA classes and LVEF-based heart failure phenotypes.
Material and Methods: This cross-sectional observational study included 57 patients. In 56 patients, LVEF phenotypes were defined as HFpEF (≥50%; n=8), HFmrEF (41–49%; n=13), and HFrEF (≤40%; n=35). Groups were compared using Kruskal–Wallis and Mann–Whitney post hoc tests. Associations were assessed using Spearman’s correlation. Holm correction was applied to post hoc comparisons and correlations. LVEF-phenotype differences were analyzed using age- and sex-adjusted multiple linear regression (OLS; HC3); NT-proBNP was log-transformed.
Results: Median (IQR) concentrations were NT-proBNP 1751 (770–10,166) pg/mL, vitamin D 44 (37–57) nmol/L, and hsCRP 11.0 (9.05–11.25) mg/L. NT-proBNP differed across NYHA classes (p=0.003). Across LVEF phenotypes, differences were observed for NT-proBNP (p=0.011) and vitamin D (p=0.027), but not hsCRP. After age/sex adjustment, the phenotype effect was not significant for NT-proBNP (p=0.077) but remained significant for vitamin D (p=0.030). HFrEF had lower adjusted vitamin D than HFmrEF (Holm p=0.033). After Holm correction, NT-proBNP correlated with NYHA (ρ=0.430; p=0.004) and LVEF (ρ=−0.481; p=0.001), and vitamin D with LVEF (ρ=0.339; p=0.042).
Conclusion: Results suggest an association of NT-proBNP with functional severity and continuous LVEF, while differences across LVEF phenotypes are attenuated after adjustment. The vitamin D–LVEF phenotype association is exploratory. Cross-sectional design, small sample size (n=57; HFpEF n=8), and possible residual confounding limit interpretation